Legal claims defining the scope of protection. Each claim is shown in both the original legal language and a plain English translation.
1. A method for video decoding in a decoder, comprising: decoding encoded video content from an encoding system, wherein the encoding system performs a scene change detection process that includes identifying a first picture among a sequence of pictures that corresponds to a last picture of a first scene, identifying a second picture among the sequence of pictures that corresponds to a first picture of a second scene, and determining that one or more pictures that are between the first picture and the second picture in the sequence of pictures correspond to an input transition scene between the first scene and the second scene, and the received encoded video content is encoded without encoding all of the pictures of the input transition scene, and includes an indication of an existence of the input transition scene and information for reproducing a transition scene that corresponds to the input transition scene; generating the transition scene that corresponds to the input transition scene based on the received indication of the existence of the input transition scene and the information for reproducing the transition scene; and outputting video content that includes the transition scene that corresponds to the input transition scene in a sequence that is between the first scene and the second scene.
Video decoding involves processing encoded video content to reconstruct scenes for display. A challenge in video encoding is efficiently handling scene transitions, where intermediate frames (input transition scenes) between two distinct scenes may not be fully encoded to save bandwidth. This invention addresses this by enabling a decoder to reconstruct transition scenes from partial data. The method begins with decoding encoded video content that was processed by an encoding system. The encoding system first identifies a first picture in a sequence as the last frame of a first scene and a second picture as the first frame of a second scene. Any frames between these two pictures are classified as an input transition scene. Instead of encoding all transition frames, the encoding system includes metadata indicating the transition's existence and data for reproducing a simplified transition scene. The decoder uses this metadata to generate a transition scene that approximates the original input transition scene. The generated transition is then inserted between the first and second scenes in the output video, ensuring smooth playback without requiring full encoding of intermediate frames. This approach reduces encoding complexity and bandwidth usage while maintaining visual continuity.
2. The method according to claim 1 , further comprising generating a cross-fade transition scene, for reproducing the transition scene that corresponds to the input transition scene, based on the information for reproducing the transition scene indicating a predetermined fading model.
This invention relates to video editing and transition effects, specifically improving the generation of cross-fade transitions between video scenes. The problem addressed is the need for automated, high-quality transition effects that smoothly blend one scene into another while maintaining visual coherence and adhering to predefined aesthetic models. The method involves generating a cross-fade transition scene that reproduces a transition effect corresponding to an input transition scene. The transition is based on information indicating a predetermined fading model, which defines parameters such as fade duration, opacity curves, and blending algorithms. The system ensures that the transition adheres to the specified model, producing a seamless and visually pleasing effect. This approach automates the transition process, reducing manual editing effort while maintaining consistency with desired visual styles. The method may also include analyzing the input scenes to determine optimal transition points, such as matching color gradients or motion trajectories, to enhance the natural flow between scenes. The fading model can be adjusted dynamically based on scene content, ensuring adaptability to different video contexts. The result is a transition that appears smooth and intentional, improving the overall viewing experience. This technique is particularly useful in automated video editing systems, where consistent and high-quality transitions are required without extensive manual intervention.
3. The method according to claim 2 , wherein a time instance for the first picture is t=0, a time instance for the second picture is t=n, the input transition scene occupies n−1 time instances from t=1 to t=n−1, each image F[t] within the cross-fade transition scene is defined as: F[t]=A*(1−s[t])+K*s[t], where s[t] is: s [ t ] = { 0 , t = 0 t / n , 0 < t < n 1 , t = n , wherein “A” is the first frame and “K” is the second frame.
This invention relates to video processing, specifically techniques for generating smooth transitions between two images or frames in a video sequence. The problem addressed is the need for a seamless and visually pleasing transition between a first image (A) and a second image (K) over a defined duration. The solution involves a cross-fade transition that interpolates between the two images over a series of frames, ensuring a gradual and controlled transition. The method defines a transition scene spanning multiple time instances, where the first image (A) is displayed at time t=0 and the second image (K) is displayed at time t=n. The transition occurs over n−1 intermediate time instances, from t=1 to t=n−1. Each intermediate frame (F[t]) in the transition is computed using a weighted combination of the first and second images, where the weight is determined by a function s[t]. The function s[t] is defined as 0 at t=0, linearly increases from 0 to 1 between t=1 and t=n−1, and reaches 1 at t=n. The intermediate frame F[t] is calculated as F[t] = A*(1−s[t]) + K*s[t], ensuring a smooth fade from the first image to the second image over the transition period. This approach provides a mathematically precise and visually smooth transition between two images in a video sequence.
4. The method according to claim 2 , wherein a time instance for the first picture is t=0, a time instance for the second picture is t=n, the input transition scene occupies n−1 time instances from t=1 to t=n−1, each image F[t] within the cross-fade transition scene is defined as: F[t]=A*(1−s[t])+K*s[t], where s[t] is: s [ t ] = { 0 , t = 0 ( t / n ) 2 , 0 < t < n 1 , t = n , wherein “A” is the first picture and “K” is the second picture.
This invention relates to digital image processing, specifically techniques for generating smooth transitions between two static images (A and K) using a cross-fade effect. The problem addressed is creating a visually pleasing transition that avoids abrupt changes while maintaining temporal coherence. The method defines a transition scene spanning n time instances, where the first image (A) appears at t=0 and the second image (K) appears at t=n. The intermediate frames (t=1 to t=n−1) are computed using a weighted blend of A and K, controlled by a smooth interpolation function s[t]. The function s[t] is zero at t=0, linearly increases as a quadratic function of t/n for 0 < t < n, and reaches 1 at t=n. Each frame F[t] in the transition is calculated as F[t] = A*(1−s[t]) + K*s[t], ensuring a gradual, mathematically smooth transition between the two images. This approach provides a predictable, aesthetically pleasing cross-fade effect without abrupt jumps or artifacts. The quadratic interpolation of s[t] ensures the transition accelerates and decelerates smoothly, enhancing visual comfort. The method is particularly useful in video editing, animation, and graphical user interfaces where seamless transitions between static images are desired.
5. The method according to claim 2 , wherein a time instance for the first picture is t=0, a time instance for the second picture is t=n, the input transition scene occupies n−1 time instances from t=1 to t=n−1, each image F[t] within the cross-fade transition scene is defined as: F[t]=A*(1−s[t])+K*s[t], where s[t] is: s [ t ] = { 0 , t = 0 ( t / n ) 1 / 2 , 0 < t < n 1 , t = n , wherein “A” is the first picture and “K” is the second picture.
This invention relates to digital image processing, specifically techniques for generating smooth transitions between two images. The problem addressed is creating a visually pleasing cross-fade transition that avoids abrupt changes while maintaining temporal coherence. The method involves defining a transition scene between a first image (A) and a second image (K) over a sequence of time instances. The transition begins at time t=0 with the first image and ends at time t=n with the second image, spanning n-1 intermediate time instances. Each intermediate image F[t] in the transition is computed as a weighted combination of the two images, where the weight is determined by a smooth transition function s[t]. For t=0, s[t] is 0, fully displaying the first image. For t=n, s[t] is 1, fully displaying the second image. For intermediate times (0 < t < n), s[t] is defined as (t/n)^(1/2), creating a non-linear fade that accelerates smoothly over time. This approach ensures a gradual, mathematically defined transition that avoids abrupt shifts in visual content. The method is particularly useful in video editing, animation, and graphical user interfaces where smooth transitions are desired.
6. The method according to claim 2 , wherein the predetermined fading model is based on a predetermined animated transition.
This invention relates to digital animation and visual effects, specifically improving the rendering of animated transitions in graphical user interfaces or multimedia applications. The problem addressed is the need for smooth, visually appealing transitions between different graphical states or elements, particularly when fading effects are involved. Existing methods often lack dynamic control over fading behavior, resulting in unnatural or inconsistent visual effects. The method involves applying a predetermined fading model to control the opacity or visibility of graphical elements during an animated transition. The fading model is not static but is instead based on a predetermined animated transition, allowing for dynamic adjustments in opacity over time. This ensures that the fading effect aligns with the animation's motion, enhancing visual coherence. The predetermined animated transition may include parameters such as duration, easing functions, or path-based motion, which influence how the fading model is applied. By linking the fading model to the animation, the method achieves more natural and synchronized visual effects compared to traditional linear or fixed-rate fading techniques. This approach is particularly useful in applications requiring high-quality animations, such as video editing software, mobile interfaces, or interactive media.
7. The method according to claim 6 , wherein the predetermined animated transition is one of a morph animation, a fade animation, a push animation, a wipe animation, a split animation, a reveal animation, a shape animation, and an uncover animation.
This invention relates to animated transitions in digital interfaces, specifically methods for dynamically altering the visual presentation of content to enhance user experience. The problem addressed is the need for smooth, engaging transitions between different visual states in graphical user interfaces (GUIs), ensuring clarity and reducing cognitive load while maintaining aesthetic appeal. The method involves applying a predetermined animated transition when switching between visual states in a GUI. These transitions include morph animations, where one element seamlessly transforms into another; fade animations, which gradually adjust opacity; push animations, where new content appears to push existing content out of view; wipe animations, which reveal new content by sweeping across the screen; split animations, where content divides and moves apart; reveal animations, which uncover new content from behind existing elements; shape animations, where content reshapes dynamically; and uncover animations, which remove covering elements to display underlying content. These transitions are triggered by user interactions or system events, ensuring intuitive and visually cohesive navigation. The method improves user engagement by making transitions more dynamic and less disruptive, while also providing developers with a standardized set of animation options to enhance interface design.
8. The method according to claim 7 , wherein the predetermined animated transition is determined at the encoding system based on the input transition scene being compared to a plurality of pattern-generated content that are each generated based respectively on at least one of the predetermined animated transitions, and the predetermined animated transition is based on the pattern-generated content which has a highest similarity to the input transition scene.
This invention relates to video encoding systems that automatically determine animated transitions between scenes. The problem addressed is the manual effort required to design and apply transitions in video production, which is time-consuming and lacks consistency. The solution involves an encoding system that analyzes an input transition scene and compares it to a library of pattern-generated content. Each pattern in the library is created using different predetermined animated transitions. The system identifies the transition pattern that most closely matches the input scene and selects the corresponding animated transition. This ensures that the chosen transition is visually coherent with the input content, improving production efficiency and quality. The method involves generating multiple pattern examples from available transitions, comparing these patterns to the input scene, and selecting the best match based on similarity metrics. This automated approach reduces the need for manual intervention while maintaining high-quality transitions.
9. The method according to claim 8 , wherein the predetermined animated transition is determined at the encoding system based on an assessment of distortion between the input transition scene and each of the plurality of pattern-generated content.
This invention relates to video encoding systems that generate animated transitions between scenes. The problem addressed is efficiently selecting an optimal animated transition that minimizes visual distortion while maintaining smooth playback. The system analyzes an input transition scene and compares it to multiple pattern-generated content options. The encoding system evaluates distortion metrics between the input scene and each candidate transition pattern. Based on this assessment, the system determines the most suitable animated transition to apply. The distortion assessment may involve analyzing pixel-level differences, motion vectors, or other visual quality metrics. The goal is to automate the selection of transitions that preserve visual coherence and reduce artifacts during playback. This approach improves encoding efficiency by avoiding manual selection or overly complex computational methods. The system may also adapt the transition based on content characteristics, such as scene complexity or motion intensity, to further optimize quality. The invention is particularly useful in video streaming, editing, and production workflows where seamless transitions are critical.
10. The method according to claim 6 , further comprising receiving an indication of the predetermined animated transition used to replace the input transition scene, along with the encoded video content, as the information for reproducing the transition scene that corresponds to the input transition scene.
This invention relates to video processing, specifically methods for encoding and reproducing animated transitions between video scenes. The problem addressed is the need to efficiently encode and reproduce dynamic transitions between video segments while maintaining high-quality visual effects. The method involves encoding video content that includes a transition scene, where the transition scene is replaced with a predetermined animated transition during playback. The encoded video content includes information for reproducing the transition scene, which corresponds to the input transition scene. Additionally, the method receives an indication of the predetermined animated transition used to replace the input transition scene, along with the encoded video content. This allows for flexible and customizable transitions during playback, improving the visual experience without requiring additional storage or bandwidth for the original transition scene. The technique ensures smooth and visually appealing transitions while optimizing encoding efficiency.
11. The method according to claim 1 , wherein the scene change detection process includes determining that the one or more picture that are between the first picture and the second picture in the sequence of picture correspond to an input transition scene between the first scene and the second scene when a duration of the one or more pictures is less than a predetermined threshold.
This invention relates to video processing, specifically scene change detection in a sequence of pictures. The problem addressed is accurately identifying transitions between scenes in a video, particularly when intermediate frames (pictures) represent a transition effect rather than a distinct scene. The method determines whether one or more pictures between a first picture (representing a first scene) and a second picture (representing a second scene) correspond to an input transition scene. This is done by analyzing the duration of the intermediate pictures. If the duration is less than a predetermined threshold, the intermediate pictures are classified as a transition scene rather than a separate scene. This helps distinguish intentional transition effects (e.g., cuts, fades) from actual scene changes, improving video editing, compression, and analysis. The method may also involve comparing visual content or other features of the intermediate pictures to confirm the transition. The threshold duration is set based on typical transition lengths in video content, ensuring accurate classification. This approach enhances scene segmentation by reducing false positives in scene change detection.
12. The method according to claim 1 , wherein the scene change detection process includes determining that the one or more pictures that are between the first picture and the second picture in the sequence of frames correspond to an input transition scene when a duration of the first scene and the second scene are each greater than a predetermined threshold.
This invention relates to video processing, specifically methods for detecting scene changes in a sequence of video frames. The problem addressed is accurately identifying transitions between distinct scenes in a video while avoiding false detections caused by brief visual changes or intermediate frames that do not represent true scene boundaries. The method involves analyzing a sequence of video frames to detect transitions between a first scene and a second scene. A scene change is confirmed only when the duration of both the first and second scenes exceeds a predetermined threshold, ensuring that brief or transient visual changes are not mistakenly classified as scene transitions. Intermediate frames between the first and second scenes are evaluated to determine if they correspond to an input transition scene, which may include effects like fades, wipes, or other transitional elements. This approach improves the reliability of scene detection by distinguishing between true scene changes and temporary visual effects. The method may also include additional steps such as analyzing frame differences, motion vectors, or other visual features to refine the detection process. By enforcing duration thresholds, the system avoids misclassifying short-lived visual changes as scene boundaries, enhancing the accuracy of scene segmentation in video processing applications. This technique is useful in video editing, content analysis, and automated video indexing systems.
13. The method according to claim 1 , wherein the indication of the existence of the input transition scene includes: a signal indicating that the second frame is the first image of the second scene along with an indication of a duration of the input transition scene.
This invention relates to video processing, specifically detecting and handling transitions between scenes in a video sequence. The problem addressed is accurately identifying the start of a new scene in a video, particularly when transitions occur, and determining the duration of the transition period. The method involves analyzing a video sequence composed of multiple frames to detect transitions between scenes. When a transition from a first scene to a second scene is detected, the system generates an indication that the second frame is the first image of the second scene. Additionally, the system provides an indication of the duration of the transition scene, which represents the time taken for the transition from the first scene to the second scene. This allows for precise scene segmentation and transition handling in video processing applications. The method may also include detecting the transition by comparing frames in the video sequence to identify changes in visual content, such as cuts, fades, or other transition effects. The duration of the transition is calculated based on the number of frames involved in the transition and the frame rate of the video. This information is useful for video editing, compression, and analysis tasks where accurate scene detection is required. The system ensures that the transition duration is accurately measured, enabling better video segmentation and processing.
14. The method according to claim 13 , wherein the indication of the duration of the input transition scene is based on a difference between a picture order count (POC) between the second picture and the first picture.
Video encoding and decoding systems often require efficient handling of scene transitions to maintain smooth playback and reduce computational overhead. A key challenge is accurately determining the duration of input transition scenes, which involves identifying the temporal relationship between consecutive video frames. Traditional methods may rely on arbitrary or fixed durations, leading to inefficiencies in encoding and potential artifacts in playback. This invention addresses the problem by dynamically calculating the duration of an input transition scene based on the difference in picture order count (POC) between two consecutive frames. The POC is a timestamp-like value assigned to video frames to indicate their display order, which is particularly useful in scenarios involving frame reordering, such as B-frames in predictive coding. By leveraging the POC difference, the method ensures precise synchronization between the input transition scene and the encoded output, improving compression efficiency and reducing latency. The approach is applicable to both encoding and decoding processes, ensuring consistent behavior across different video processing stages. This method enhances the accuracy of scene transition handling in video compression standards like H.264, H.265, and beyond.
15. The method according to claim 13 , wherein the indication of the duration of the input transition scene is based on a duration time value being included in the header of the second picture.
This invention relates to video processing, specifically methods for handling input transition scenes between video segments. The problem addressed is the need to accurately determine the duration of transition scenes, such as fades, wipes, or other effects, to ensure smooth playback and proper synchronization in video editing or streaming applications. The method involves analyzing a second picture in a sequence of pictures to detect an input transition scene. The duration of this transition is determined by extracting a duration time value embedded in the header of the second picture. This value is then used to control the playback or processing of the transition scene, ensuring it is displayed for the correct amount of time. The method may also involve adjusting the duration based on additional factors, such as the type of transition or user preferences, to enhance visual quality or performance. The invention is particularly useful in video editing software, streaming platforms, or playback devices where precise timing of transitions is critical. By embedding the duration directly in the picture header, the method simplifies the detection and processing of transitions, reducing computational overhead and improving efficiency. This approach ensures consistent playback across different devices and platforms, enhancing the overall viewing experience.
16. The method according to claim 13 , wherein the signal is included in a header of the second picture.
A method for processing video data involves encoding and transmitting picture data between devices. The method addresses the challenge of efficiently conveying information about picture dependencies in video encoding, particularly in systems where pictures may be encoded independently or with references to other pictures. The method includes generating a signal indicating whether a second picture is encoded independently or depends on a first picture. This signal is included in a header of the second picture, allowing a decoder to determine the encoding mode without additional processing. The method ensures compatibility with existing video coding standards while improving decoding efficiency by reducing the need for redundant signaling. The signal may be a flag or a syntax element that explicitly states the encoding relationship, enabling decoders to quickly identify whether the second picture is self-contained or relies on prior data. This approach optimizes bandwidth usage and processing time by avoiding unnecessary reference checks, particularly in scenarios where pictures are transmitted out of order or independently. The method is applicable to video streaming, broadcasting, and real-time communication systems where efficient picture handling is critical.
17. The method according to claim 13 , wherein the signal is included a Supplementary Enhancement Information (SEI) message.
This invention relates to video encoding and decoding, specifically improving the transmission of video data by embedding additional information in a Supplementary Enhancement Information (SEI) message. The problem addressed is the need for efficient and flexible transmission of metadata or auxiliary data alongside video streams without disrupting the core video decoding process. SEI messages are used in video coding standards like H.264 and H.265 to carry non-essential but useful data, such as timing information, region-of-interest data, or other enhancements. The invention enhances this by including a specific signal within an SEI message, allowing for better synchronization, error resilience, or other improvements in video processing. The method involves generating or processing the SEI message to include this signal, which can then be extracted and utilized by a decoder to improve video playback, error handling, or other functionalities. This approach ensures that the additional data does not interfere with the mandatory video decoding process while providing valuable enhancements. The invention is particularly useful in applications requiring robust metadata transmission, such as adaptive streaming, video conferencing, or broadcast systems.
18. The method according to claim 1 , wherein the second picture is decoded prior to the generation of the transition scene that corresponds to the input transition scene, and the transition scene that corresponds to the input transition scene is output prior to the second scene being outputted in the outputted video content.
This invention relates to video processing, specifically methods for generating and outputting transition scenes between video segments. The problem addressed is the need for efficient and seamless transitions between video scenes, particularly in real-time or near-real-time applications where processing delays can disrupt the viewing experience. The method involves decoding a second picture (e.g., a video frame or scene) before generating a transition scene that corresponds to an input transition scene. The transition scene is then outputted before the second scene is displayed in the final video content. This ensures that the transition is ready for display at the correct time, avoiding delays or interruptions. The process may include additional steps such as encoding or decoding other video segments, synchronizing audio, or adjusting timing to maintain smooth playback. By decoding the second picture in advance, the method reduces latency and ensures that the transition scene is available when needed, improving the overall quality of the video output. This approach is particularly useful in applications like live streaming, video editing, or adaptive bitrate streaming, where minimizing delays is critical. The invention may also include techniques for handling different types of transitions, such as cuts, fades, or wipes, while maintaining synchronization with audio and other multimedia elements.
19. An apparatus, comprising: processing circuitry configured to: decode encoded video content from an encoding system, wherein the encoding system performs a scene change detection process that includes identifying a first picture among a sequence of pictures that corresponds to a last picture of a first scene, identifying a second picture among the sequence of pictures that corresponds to a first picture of a second scene, and determining that one or more pictures that are between the first picture and the second picture in the sequence of pictures correspond to an input transition scene between the first scene and the second scene, and the received encoded video content is encoded without encoding all of the pictures of the input transition scene, and includes an indication of an existence of the input transition scene and information for reproducing a transition scene that corresponds to the input transition scene; generate the transition scene that corresponds to the input transition scene based on the received indication of the existence of the input transition scene and the information for reproducing the transition scene; and output video content that includes the transition scene that corresponds to the input transition scene in a sequence that is between the first scene and the second scene.
The apparatus is designed for video processing, specifically for handling encoded video content that includes scene transitions. The problem addressed is the inefficient encoding and transmission of transition scenes between distinct video scenes, which can consume significant bandwidth and processing resources. The apparatus receives encoded video content from an encoding system that has already performed scene change detection. This detection process identifies the last picture of a first scene (first picture) and the first picture of a subsequent second scene (second picture). The encoding system then determines that the pictures between these two key frames represent an input transition scene. Instead of encoding all pictures in this transition scene, the encoding system includes an indication of the transition scene's existence and metadata for reproducing it. The apparatus decodes this encoded content, generates the transition scene based on the provided indication and metadata, and outputs video content where the generated transition scene is inserted between the first and second scenes. This approach reduces bandwidth and processing requirements by avoiding full encoding of transition scenes while maintaining visual continuity.
20. A non-transitory computer readable medium storing instructions which when executed by a computer for video decoding cause the computer to perform: decoding encoded video content from an encoding system, wherein the encoding system performs a scene change detection process that includes identifying a first picture among a sequence of pictures that corresponds to a last picture of a first scene, identifying a second picture among the sequence of pictures that corresponds to a first picture of a second scene, and determining that one or more pictures that are between the first picture and the second picture in the sequence of pictures correspond to an input transition scene between the first scene and the second scene, and the received encoded video content is encoded without encoding all of the pictures of the input transition scene, and includes an indication of an existence of the input transition scene and information for reproducing a transition scene that corresponds to the input transition scene; generating the transition scene that corresponds to the input transition scene based on the received indication of the existence of the input transition scene and the information for reproducing the transition scene; and outputting video content that includes the transition scene that corresponds to the input transition scene in a sequence that is between the first scene and the second scene.
Video encoding and decoding systems often struggle with efficiently handling scene transitions, which can lead to unnecessary data storage and processing overhead. During scene changes, intermediate frames (transition scenes) between two distinct scenes are typically encoded, increasing file size and computational load without significantly improving visual quality. This invention addresses the problem by selectively encoding only key frames of the transition scene while providing metadata to reconstruct the transition effect during decoding. The system first identifies the last frame of a first scene and the first frame of a subsequent second scene. Any intermediate frames between these key frames are classified as a transition scene. Instead of encoding all transition frames, the encoding system generates metadata describing the transition effect (e.g., fade, wipe, dissolve) and includes this metadata in the encoded output. The decoder then reconstructs the transition scene using the metadata, inserting it between the first and second scenes during playback. This approach reduces storage and bandwidth requirements while maintaining smooth scene transitions. The solution is particularly useful in video streaming, editing, and archival applications where efficiency is critical.
Unknown
April 28, 2020
Browse 5M+ US patents with plain-English claim translations and AI-generated analysis.